Join GitHub today
Extend Dataset API to support writing to files (not just reading) #15014
Just as we have a pipeline to read from files and convert them to output vectors/data via some model, It would be very useful to have a pipeline to take data and write it to files.
Otherwise, we have to use TFRecordWriters and a bit more code, but really much of this can be abstracted away, simplifying the data creation process, especially when writing to shards (as one might do using distributed Tensorflow).
Thank you for your post. We noticed you have not filled out the following field in the issue template. Could you update them if they are relevant in your case, or leave them as N/A? Thanks.